BytesAgain is a curated directory of 60,000+ AI agent skills from ClawHub, GitHub, LobeHub, and Dify. Search skills by keyword in 7 languages, browse by role (developer, creator, trader, marketer) or by use case.

How do I find AI skills on BytesAgain?

Use the search bar on BytesAgain.com to search by keyword in 7 languages. You can also browse by role (developer, creator, trader, marketer) or by use case. Each skill shows install instructions for Claude, Cursor, OpenClaw, Continue, and more.

Yes, BytesAgain is completely free. No registration required for searching skills. The MCP API is also free with rate limits.

Does BytesAgain have an API for AI agents?

Yes! BytesAgain provides a free MCP SSE endpoint at /api/mcp/sse for AI agents, plus a REST API at /api/mcp?action=search&q= . No authentication needed.

Can I request a new AI skill on BytesAgain?

Yes! Visit the Requests page on BytesAgain.com to submit a skill request. Your request will be visible to the community and notified to the site admin.

🦀 ClawHub

The Librarian

by @rochyroch

Build and search lightweight quantized document indexes with TurboVec. Use when you need to create searchable indexes from documents for RAG applications wit...

Versionv1.0.1

Comments1

#crypto-defi #legal

⚡ When to Use

Trigger	Action
\|----------\|---------------------\|
\| Resource-constrained hardware \| ✅ Runs on Raspberry Pi, 512MB RAM \|
\| Personal knowledge base \| ✅ Zero infrastructure \|
\| Embedded/offline deployment \| ✅ No cloud, no database \|
\| 100K+ documents on limited hardware \| ✅ Fits where FAISS doesn't \|
\| Medical/legal records \| ❌ Use FAISS instead \|
\| Maximum accuracy required \| ❌ Use FAISS + Flashrank \|
Accuracy: ~97-98% of FAISS for 4-bit quantization. Top results may occasionally swap ranking.

💡 Examples

Prerequisites

# Install BLAS library (required for TurboVec)
sudo apt install libblas3
Create venv and install dependencies
cd /path/to/the-librarian
python3 -m venv venv
source venv/bin/activate
pip install turbovec numpy requests rank-bm25 flashrank

Build an Index

# Using the wrapper (recommended)
./scripts/librarian build /path/to/documents/ index/my_library
With options
./scripts/librarian build /path/to/docs/ index/my_library --bits 3 --chunk-size 800
Direct Python
LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libblas.so.3 \
  python scripts/build_index.py --input /path/to/docs/ --output index/my_library

Search

# Pure vector search
./scripts/librarian search "habit formation" index/my_library
Hybrid (vector + BM25)
./scripts/librarian search "habit formation" index/my_library --hybrid
Hybrid + rerank (best accuracy)
./scripts/librarian search "habit formation" index/my_library --hybrid --rerank
With context expansion
./scripts/librarian search "habit formation" index/my_library --hybrid --rerank --expand 1
JSON output
./scripts/librarian search "habit formation" index/my_library --json

⚙️ Configuration

# Install BLAS library (required for TurboVec)
sudo apt install libblas3
Create venv and install dependencies
cd /path/to/the-librarian
python3 -m venv venv
source venv/bin/activate
pip install turbovec numpy requests rank-bm25 flashrank

Build an Index

# Using the wrapper (recommended)
./scripts/librarian build /path/to/documents/ index/my_library
With options
./scripts/librarian build /path/to/docs/ index/my_library --bits 3 --chunk-size 800
Direct Python
LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libblas.so.3 \
  python scripts/build_index.py --input /path/to/docs/ --output index/my_library

Search

# Pure vector search
./scripts/librarian search "habit formation" index/my_library
Hybrid (vector + BM25)
./scripts/librarian search "habit formation" index/my_library --hybrid
Hybrid + rerank (best accuracy)
./scripts/librarian search "habit formation" index/my_library --hybrid --rerank
With context expansion
./scripts/librarian search "habit formation" index/my_library --hybrid --rerank --expand 1
JSON output
./scripts/librarian search "habit formation" index/my_library --json

View on ClawHub

TERMINAL

clawhub install thelibrarian

🧪 Use this skill with your agent

Most visitors already have an agent. Pick your environment, install or copy the workflow, then run the smoke-test prompt above.

Manusinvite

Task-oriented agent. Great for testing AI skills end-to-end.

Try Manus →

OpenClaw

Local-first agent. Install skills via ClawHub CLI.

Set up OpenClaw →

Claude Code

Anthropic's coding agent. Paste the prompt or SKILL.md into your session.

Claude Code docs →

Cursor

AI-powered IDE. Use the smoke-test prompt in Cursor Agent.

Open Cursor →

Continue.dev

Open-source AI code assistant. Add SKILL.md as a custom tool.

Continue docs →

Windsurf

Agentic IDE by Codeium. Paste the prompt into Cascade.

Try Windsurf →

Cline

VS Code extension for autonomous coding with MCP tools.

Cline on GitHub →

Copilot Workspace

GitHub's AI dev environment. Suitable for code-generation skills.

Copilot Workspace →

Trigger	Action
\|----------\|---------------------\|
\| Resource-constrained hardware \| ✅ Runs on Raspberry Pi, 512MB RAM \|
\| Personal knowledge base \| ✅ Zero infrastructure \|
\| Embedded/offline deployment \| ✅ No cloud, no database \|
\| 100K+ documents on limited hardware \| ✅ Fits where FAISS doesn't \|
\| Medical/legal records \| ❌ Use FAISS instead \|
\| Maximum accuracy required \| ❌ Use FAISS + Flashrank \|
Accuracy: ~97-98% of FAISS for 4-bit quantization. Top results may occasionally swap ranking.